In my new Dissection column over at Wired, I take a look at a remarkable new experiment on E. coli. Scientists randomly rewired the network of genes that control much of the microbe's activity and found that it generally just kept humming along.
One thing worth adding...in an accompanying commentary, Matthew Bennett and Jeff Hasty at UCSD write,
This conclusion also flies in the face of the popular misconception among opponents of the evolutionary theory, who believe that the genetic code is irreducibly complex. For instance, advocates of 'intelligent design' compare the genome to modern engineered machines such as integrated circuits and clocks, which will cease to function if their internal design is altered. Although sometimes it is instructive to point to similarities between the design principles behind modern technology and those behind genetics, the analogy can only go so far. Engineered devices are generally designed to work just above the point of failure, so that any tampering with their construction will result in catastrophe. In the event of failure, new clocks can be purchased or central processing units replaced. But nature does not have that option. To survive -- and so evolve -- organisms must be able to tolerate random mutations, deletions and recombination events. And Isalan and colleagues' work provides an important step forward in quantifying just how robust the genetic code can be.
Another reason why I ended up writing a whole book about this bug.
Comments
Escherichia rules!
Posted by: 6EQUJ5 | April 19, 2008 8:18 AM
As an engineer, the irreducible complexity argument seems wrong on another level. New products are not generally designed in a vacuum - they build on the successes of previous gadgets. So, Watt's steam engine wasn't the first, it was just so much better than older engines that it was practical for many new uses. And the internal combustion engine builds on steam engine tech. And little bits of these things, valves and such, can often be removed and used for unrelated purposes.
And, all the same, signs of irreducible complexity in the design of life wouldn't be evidence that God did it. It doesn't say much either way. Genetic algorithms are really pretty good at these things. If anything, i wonder less that genetics could make animals like humans, and more that it took so long. That suggests that the goal was probably something else entirely.
Posted by: Stephen | April 23, 2008 1:06 PM
Or that there was no goal at all.
Posted by: Ray Freeman-Lynde | April 26, 2008 5:32 PM